Despite significant advances in identifying genes implicated in ASD, the neural circuit mechanisms that contribute to impaired social behaviors and communication in ASD remain elusive. This lack of knowledge represents a critical gap in the development of circuity-based treatment. Social interactions demand complex neural computations, including sensory processing of social cues, decisions to determine appropriate behavioral responses, and planning and execution of the motor programs necessary to enact these behaviors. Using neuroimaging studies, the general structure of neural networks involving amygdala, hypothalamus, thalamus, ventral tegmental area, nucleus accumbens, and ventral hippocampus have been associated with ?social circuitry?. However, their exact neuron ensembles and how they mediate social behaviors remains poorly defined. Our overarching hypothesis is that the functional connectivity of these circuits is altered in ASD. The neural process responsible for social behaviors most likely results from the emergent properties of transiently active neural ensembles in social behavior circuits. In this application, we propose, for the first time, to use novel and innovative tools that enable neurons in these ensembles to be permanently tagged and subsequently manipulated in the living animal. Combining recent advances of Dr. Yong-hui Jiang (PI) and Dr. Fan Wang?s (Co-investigator) groups creates a unique opportunity to explore this direction. Jiang?s group recently produced an autism model with Shank3 complete deficiency by deleting exon 4-22 (?e4-22) that has strong ?construct? and ?face? validity for SHANK3-related ASD. Shank3?e4-22-/- mice recapitulate the ASD-like behaviors with impairments in social interaction and communication, as well as aberrant functional connectivity. Wang?s group developed a highly innovative technique: Capturing Activated Neuronal Ensembles (CANE). This novel technique ?captures? and ?manipulates? neuronal ensembles in mouse brains during behavior interaction. Combining the best autism model and innovative techniques provides an unprecedented opportunity to explore the most important question in modeling autism. We hypothesize that the complete deficiency of Shank3 leads to altered neural ensembles in social circuitry that underlie the observed impaired social behaviors. Our objective is to use the CANE method to identify the neural ensembles that underlie autism behavior in Shank3 mouse model. The causality between impaired circuit and behaviors will be investigated by in vivo recording and optogenetic manipulation. Our study is first application of the CANE method to dissect the social circuity in a genetically modified ASD mouse model and represents the first step toward the development of circuit specific treatment for ASD. More importantly, the success of this project will support a paradigm shift in how we model autism, as well as delineate the circuitry for other behaviors.